Geography Reference
In-Depth Information
been treated by economic geographers in a truly evolutionary manner. Broadly speak-
ing, they came to four lines of criticism: (1) institutions are often presented as pre-given
and i xed, as if these come from nowhere (a-historical) and do not change over time; (2)
institutions are often depicted as factors that determine, instead of condition, the eco-
nomic behaviour of agents and the performance of regions; (3) institutional approaches
in economic geography tend to employ case-study approaches while ignoring, if not
rejecting altogether, the use of quantitative methodologies. Although case studies have
generated many valuable insights into processes of regional innovation, many institu-
tional approaches have been reluctant to test any hypotheses that might be derived from
these; (4) institutional approaches in economic geography tend to associate institutions
with territories (at whatever spatial level) and, hence, spatial dif erences in economic
activities are attributed to institutional dif erences among territories. This stands in
contrast to those evolutionary approaches to economic geography that reason from
organisational routines, and that view the behaviour of i rms as mainly stemming from
their routines, rather than from territorial institutions.
When taking such a i rm-based, micro-perspective, one avoids running the risk of
over-emphasising the role of territorial institutions and violating a crucial ingredient of
an evolutionary approach, that is, the heterogeneity of i rms. The assumption that all
agents act or perform the same when subject to the same regional institutions also con-
tradicts empirical i ndings that suggest the opposite.
As noticed above, Giuliani (2007),
among others, has demonstrated that agents in clusters dif er widely in terms of eco-
nomic power, absorptive capacity and network position, despite the fact that clusters are
associated with a particular set of institutions. This variety can only be understood from
the fact that i rms develop routines in a path-dependent and idiosyncratic manner, and
territorial institutions are often so general such that specii c ef ects at the i rm level can
still vary greatly (Boschma and Frenken, 2009a). The very fact that a variety of routines
is found in a territory shows that territory-specii c institutions do not determine the kinds
of routine that develop and survive in a territory. As Gertler (2009) puts it, there is 'a
danger of “reading of ” individual behaviour from territorial institutions'. Instead, there
is a need to account for the role of contingency in regional development (see e.g. Bathelt
and Glückler, 2003). This implies taking the individual and i rm level more seriously than
institutional studies have tended to in the past.
This is not to deny that territorial institutions may explain part of the inter-regional
variety of routines, however. For example, production techniques of plants in a number
of manufacturing industries have been found to be more similar within than across US
regions, and that these regional dif erences are quite persistent over time (Essletzbichler
and Rigby, 2005; Rigby and Essletzbichler, 1997). This may be attributable to region-
specii c institutions, but it may also be the result of routine replication among local i rms
through spinof dynamics or labour mobility ef ects. Therefore, we have to be cautious
about taking the impact of institutions for granted, and assess their relative importance
on a case by case basis (Boschma and Frenken, 2009a).
To incorporate institutions more fully into an evolutionary economic geography
framework, MacKinnon et al. (2009) claim there is a need to account for power and
labour-capital conl icts. Boschma and Frenken (2009a) have argued that this can be
accomplished when accounting for the political dimension of routines, as advocated by
Nelson and Winter (1982). Besides the cognitive dimension of routines, which has drawn
